Secondary Traumatic Stress, Burnout and Resilience in the Child Welfare Workforce: Early Results from Nebraska's Randomized Controlled Trial of Resilience Alliance

Background and Purpose: In 2018 a needs assessment indicated that over 50% of the child welfare workforce in Nebraska was experiencing elevated secondary traumatic stress. Thus, the state’s QID-WD team undertook a phased intervention known as “CFS Strong.” The first phase included the 24-week Resilience Alliance (RA) program (developed by the Administration for Children’s Services and NY University Children’s Trauma Institute), implemented between May and December 2019. We analyze changes in caseworker and supervisor secondary traumatic stress, burnout and resilience over the 6-month period of RA participation.

Methods: A cluster randomized controlled trial (cRCT) was employed to evaluate the effectiveness of RA. The unit of randomization was a supervisory group (supervisor + workers), with groups stratified by the five Nebraska service areas and equal assignment to intervention and control. Supervisory units assigned to the intervention were combined into groups of 10-15 employees to form RA groups. All participants in both the intervention and control groups completed the Secondary Traumatic Stress scale (Bride, et al, 2004), the Maslach Short Burnout Scale and the Connor Davidson Resilience scale prior to and at completion of the RA program. We ran linear, mixed repeated measures models in SPSS, moderated by years of child welfare work experience and level of intake-assessment experience.

Results: Randomized groups were balanced on a variety of demographic factors. Overall attendance of workers and supervisors at RA group meetings was strong, with the median number of sessions attended at 18 (75%). Mixed modeling results show no differences between intervention and control in change over time for STS, burnout or resilience. However, for some segments of workers, there may be a protective effect of RA. Among staff with two or fewer years of child welfare work experience, those in the intervention group showed flat burnout during RA, while those not participating in RA showed a significant increase in burnout (n=328; F=4.339; p=.038; α=.05). For the resilience outcome, there was a borderline significant drop in resilience among control group employees with over two years of experience (n=328; F=3.601; p=.059; α=.10), but no significant drop among those in the intervention group. Finally, there was a significant increase in resilience among staff in the intervention group with “a great deal” of intake experience (n=316; F=5.191; p=.024; α=.05), but no corresponding increase in the control group.

Conclusions and Implications: To our knowledge, this study is the first randomized controlled trial of the Resilience Alliance program. RA is currently listed on the California Evidence Based Clearinghouse as unable to be rated for evidence of efficacy because it lacks at least one "published, peer-reviewed study utilizing some form of control." Therefore, the study makes a significant contribution to the knowledge base. Although RA may not have been broadly effective across the child welfare workforce in Nebraska, it may indeed offer promise in mitigating burnout and supporting resilience in some segments of the workforce.